| *** jamesmcarthur has quit IRC | 00:15 | |
| *** tosky has quit IRC | 00:18 | |
| *** jamesmcarthur has joined #storyboard | 00:18 | |
| *** jamesmcarthur has quit IRC | 00:20 | |
| *** jamesmcarthur has joined #storyboard | 00:23 | |
| *** jamesmcarthur has quit IRC | 00:26 | |
| *** jamesmcarthur has joined #storyboard | 00:27 | |
| *** jamesmcarthur has quit IRC | 00:27 | |
| *** jamesmcarthur has joined #storyboard | 00:39 | |
| *** jamesmcarthur has quit IRC | 00:48 | |
| *** jamesmcarthur has joined #storyboard | 01:06 | |
| *** jamesmcarthur has quit IRC | 01:28 | |
| *** jamesmcarthur has joined #storyboard | 02:03 | |
| *** tellesnobrega has quit IRC | 02:14 | |
| *** tellesnobrega has joined #storyboard | 02:23 | |
| *** udesale has joined #storyboard | 04:00 | |
| *** jtomasek has joined #storyboard | 05:41 | |
| *** jamesmcarthur has quit IRC | 06:26 | |
| *** jtomasek has quit IRC | 06:52 | |
| *** jtomasek has joined #storyboard | 09:59 | |
| *** jtomasek has quit IRC | 10:00 | |
| *** jtomasek has joined #storyboard | 10:00 | |
| *** tosky has joined #storyboard | 10:03 | |
| *** udesale has quit IRC | 10:50 | |
| *** jamesmcarthur has joined #storyboard | 11:22 | |
| *** jamesmcarthur has quit IRC | 11:28 | |
| *** teor4rg has joined #storyboard | 12:29 | |
| *** udesale has joined #storyboard | 13:31 | |
| *** udesale has quit IRC | 13:46 | |
| fungi | diablo_rojo: tosky: so i'm ready to initiate the sahara imports if timing's okay... still following the mapping from http://paste.openstack.org/show/698760/ ? | 14:41 |
|---|---|---|
| tosky | fungi: yes, thanks, exactly | 14:42 |
| tosky | fungi: do I understand it correctly that the import will be reexecuted at the end to catch the new bugs/updates to the existing bugs that happened in the meantime? | 14:42 |
| tosky | we have one or two reviews in flight | 14:42 |
| fungi | given the duration the initial imports will take i'll rerun them again at the end to catch any stragglers | 14:43 |
| fungi | i'll get it underway momentarily and let you know as soon as it completes | 14:43 |
| tosky | ack, thanks; I think that after those one or two changes, from today's evening we can refrain from merging patches until Monday, when the migration can be considered as complete | 14:44 |
| tosky | tellesnobrega: ^^ | 14:44 |
| fungi | i've batched up the 7 import commands and started them now. running under time(1) so we have a good comparative benchmark | 14:51 |
| fungi | imported stories are starting to appear under https://storyboard.openstack.org/#!/project_group/74 though only if you expand the filter beyond active stories (doesn't look like the import has hit one just yet) | 14:52 |
| tosky | probably not, I don't think we have many active stuff for python-saharaclient (the first in the list) | 14:55 |
| tosky | fungi: I see https://storyboard.openstack.org/#!/story/1519510 - but the python-saharaclient task was merged, shouldn't it be hidden in the Active view for python-saharaclient? | 14:59 |
| fungi | tosky: i suspect that's a feature request for the project view page. the story itself is still considered active because it has incomplete tasks for other projects | 15:02 |
| fungi | i can think of a couple different ways we might want to slice that so as to hide stories which have complete tasks for the project(s) on which you're filtering but open tasks for other projects | 15:03 |
| fungi | right now i'm pretty sure it's just basing it on the overall story state instead (which is inferred from the set of its tasks) | 15:04 |
| tosky | I understand that the story is still open and in some views it should show, but I think that I filtering by repository, I don't need to see it | 15:04 |
| fungi | tosky: i agree, that seems like a reasonable expectation to me too | 15:05 |
| tosky | unrelated question: do you think if it's possible (or it would make sense) to get notifications from storyboard consumed by one of the bots in the channel | 15:06 |
| tosky | so that notifications of new stories can be reported on the IRC channels? | 15:07 |
| *** jamesmcarthur has joined #storyboard | 15:07 | |
| *** jamesmcarthur has quit IRC | 15:07 | |
| *** jamesmcarthur has joined #storyboard | 15:07 | |
| persia | I can confirm that the current behaviour for story filters for "active" and "project=foo" is to get all stories where at least one task is project foo (with any status) and at least one task is active. | 15:10 |
| *** jamesmca_ has joined #storyboard | 15:11 | |
| persia | An argument in support of that is that while a team has done something towards the story, the story isn't complete, and the team may need to do more to get it complete. I'm not sure that argument is as useful for larger deployments (like OpenStack), where there are likely to be stories that involve hundreds of tasks, where each task is small, and many of the projects affected are not stakeholders in the story. | 15:11 |
| *** jamesmcarthur has quit IRC | 15:11 | |
| tosky | persia: if there are no task assigned to my project I'm filtering against, I don't see why the story is still relevant for that specific view | 15:15 |
| tosky | if the team may need to do more, it means that (at least) one other open task should exist | 15:15 |
| *** udesale has joined #storyboard | 15:16 | |
| tosky | the example that we are discussing is https://storyboard.openstack.org/#!/story/1519510 visible in https://storyboard.openstack.org/#!/project/934 | 15:16 |
| persia | Yes. I believe this is because OpenStack is large. For a smaller deployment, where all participants are aligned towards the stories, the "someone else's problem" effect is smaller. That many of the deployments driving development for a while were smaller likely contributes to this not being perceived as too much of a bug. | 15:17 |
| tosky | it's not so much a "someone else's problem": for the point of view of the python-saharaclient repository, the task is closed and the story is done | 15:18 |
| persia | Whereas, as the OpenStack deployment is becoming larger (as more folk migrate), I expect addressing this would be more of a priority. | 15:18 |
| persia | tosky: It's "someone else's problem", in that the a repository-specific viewpoint is adopted, rather than a deployment-wide viewpoint. | 15:19 |
| tosky | probably in a smaller project you are monitoring the global list of stories, so you will catch that anyway somewhere else; just not in that view | 15:19 |
| persia | Precisely. | 15:19 |
| tosky | and we have this a deployment-wide viewpoint https://storyboard.openstack.org/#!/story/list | 15:19 |
| persia | And the majority of early use of the OpenStack deployment was "all the infra projects", where that global "solve the stories over many repos" was more prevalent. | 15:19 |
| tosky | do you know if there is a story to change or discuss this, or should I create one? | 15:21 |
| persia | To be extra-clear: I'm not saying "this isn't a bug", but rather saying "this is a use case that also needs supporting" + saying "that this is a bug probably wasn't noticed for $reasons". I believe the underlying cause is that it makes the filter logic easier. | 15:21 |
| fungi | yeah, i have a feeling the implementation would end up being a toggle for something like "base story status on filtered project list" | 15:22 |
| fungi | so you could switch that on and off in the query/project view | 15:22 |
| tosky | or add another tab "Related [Active]" after "Active", "Merged", "Invalid" and group those stories | 15:23 |
| fungi | however that means extending the api, i expect, since the alternative would be baking business logic into the webclient which seems like a bad idea | 15:23 |
| persia | I don't see such a story from a quick few searches of the stories affecting the storyboard group. | 15:23 |
| fungi | right, it's likely to be an increasing annoyance for teams which have shared stories with other teams | 15:24 |
| fungi | more isolated projects like storyboard itself won't generally notice | 15:24 |
| persia | Indeed. To my knowledge, this is currently the largest StoryBoard deployment, and as it grows, we'll probably notice other scaling issues. | 15:25 |
| tosky | ok, let me write something | 15:25 |
| persia | (as this is an issue for the intersection of "these stories contain lots of tasks against lots of projects" and "there are too many things going on for everyone to care about everything") | 15:25 |
| fungi | tosky: as for bot integration, i think mtreinish has/had some patches to get mqtt event publication so an irc bot could just implement an mqtt subscriber and listen to firehose.openstack.org for whatever event filter is relevant to it | 15:26 |
| tosky | oh, nice to hear | 15:26 |
| fungi | bonus is that the firehose already contains launchpad events (implemented through a roundabout smtp/pop3 solution, but it works) | 15:26 |
| persia | I think https://review.openstack.org/#/c/540958/3 is the tip of the MQTT stuff, which needs more review/testing. | 15:27 |
| fungi | so one bot could probably consume events for both | 15:27 |
| *** udesale has quit IRC | 15:44 | |
| *** jdandrea_ has joined #storyboard | 15:55 | |
| *** jamesmca_ has quit IRC | 15:58 | |
| *** jdandrea has quit IRC | 16:03 | |
| *** lifeless has quit IRC | 16:03 | |
| *** davidlenwell has quit IRC | 16:03 | |
| *** krotscheck has quit IRC | 16:03 | |
| *** lifeless has joined #storyboard | 16:05 | |
| *** davidlenwell has joined #storyboard | 16:11 | |
| *** krotscheck has joined #storyboard | 16:18 | |
| *** udesale has joined #storyboard | 17:07 | |
| *** diablo_rojo has quit IRC | 17:14 | |
| *** diablo_rojo has joined #storyboard | 17:14 | |
| *** udesale has quit IRC | 17:16 | |
| SotK | when I was testing that set of changes I ran into some strange issues and wanted to look deeper into if I was doing something stupid with my new test instance or if it was a bug introduced in those changes | 17:20 |
| SotK | I've not yet had chance to do said looking | 17:20 |
| *** jtomasek has quit IRC | 17:33 | |
| diablo_rojo | SotK, do you remember what you were doing? I can try to duplicate. | 17:34 |
| SotK | yeah, I set up notifications as if I were using rabbitmq, but using the mqtt config keys instead | 17:35 |
| SotK | then ran the worker daemons | 17:36 |
| SotK | I found that some events were being missed and others were being duplicated multiple times | 17:36 |
| * SotK will write more detailed explanation later | 17:36 | |
| *** jamesmcarthur has joined #storyboard | 18:15 | |
| *** jamesmcarthur has quit IRC | 18:20 | |
| diablo_rojo | SotK, sounds good. | 18:45 |
| *** jamesmcarthur has joined #storyboard | 19:23 | |
| *** jamesmcarthur has quit IRC | 19:26 | |
| *** jamesmcarthur has joined #storyboard | 19:28 | |
| *** jtomasek has joined #storyboard | 19:44 | |
| *** jtomasek has quit IRC | 19:45 | |
| fungi | diablo_rojo: tosky: initial import spanned 201m31.462s wall clock time. i've initiated a second catch-up run now | 20:09 |
| fungi | timing it as well to see how much faster the subsequent run is (if any) | 20:09 |
| diablo_rojo | fungi, woot woot! | 20:10 |
| diablo_rojo | Thanks for running that fungi :) | 20:10 |
| diablo_rojo | Another project migrated! | 20:10 |
| SotK | \o/ | 20:16 |
| *** jamesmcarthur has quit IRC | 20:16 | |
| diablo_rojo | Next on the docket is Ironic! | 20:18 |
| * diablo_rojo needs to convene with TripleO about what squad wants to take the plunge first. | 20:18 | |
| *** jamesmcarthur has joined #storyboard | 21:22 | |
| *** jamesmcarthur has quit IRC | 21:48 | |
| *** lifeless has quit IRC | 22:07 | |
| *** lifeless has joined #storyboard | 22:07 | |
| *** lifeless has quit IRC | 22:08 | |
| *** lifeless has joined #storyboard | 22:08 | |
| fungi | okay, so rerunning the full import for sahara to incrementally pick up any new changes to their bugs took 130m2.155s wall clock time | 22:22 |
| fungi | certainly faster than the initial import, but not by a significant factor | 22:23 |
| fungi | ~2/3 as long | 22:23 |
| diablo_rojo | Interesting. Any idea how many new things it had to import during the second run fungi ? | 22:26 |
| fungi | i can skim | 22:27 |
| fungi | no idea... i only seem to have 1805 lines of scrollback buffer in my terminal and most of that is "WARNING: Skipping bug https://api.launchpad.net/1.0/bugs/<...> due to tag rules" | 22:28 |
| persia | If binary logging is enabled on the server, one could look at the transaction journal, if one had credentials. I suspect it is not enabled. | 22:34 |
| persia | err, database server, database transaction journal. | 22:35 |
| persia | Storyboard has events, but I'm not sure there is any good way to determine which stories were created, as I think the import falsifies data to match that received from LP. | 22:36 |
| SotK | diablo_rojo: so, to start with I installed mosquitto (assuming ubuntu, `sudo apt install mosquitto` will install it and start the service) | 23:25 |
| SotK | then, add `enable_notifications = True` to the DEFAULT section of storyboard.conf (it should already be there commented out on line 42) | 23:27 |
| SotK | next, in the notifications section, add `driver = mqtt` | 23:28 |
| SotK | also add a new section called `mqtt_notifications`, and in that add `hostname = localhost` | 23:29 |
| SotK | then, run the api and also run `tox -e venv -- storyboard-worker-daemon --config-file etc/storyboard.conf` | 23:29 |
| *** openstackgerrit has quit IRC | 23:48 | |
| diablo_rojo | SotK, got it. Will give that a try and tell you what I see | 23:57 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!